Previous abstract
Next abstract
Session Q04 - DNA-Ligand Interactions
Invited session, Friday morning, March 24, 8:00
Room J3, San Jose Convention Center
One of the major targets of antitumour and antibiotic drugs is DNA. Drugs bind to DNA molecules by a variety of interactions. The strength of the binding resulting from these interactions can be measured experimentally by the DNA-drug equilibrium binding constant. This binding constant can be calculated from a study of the dissociation of a drug from a DNA-drug complex whose molecular structure can be determined from X-ray crystal measurements. We have developed a computational method, based on an effective harmonic theory, that gives a statistical estimate of the disruption probability of the individual DNA-drug chemical bonds and the probability for the disruption of the collective nonbonded interactions between DNA and the drug leading to final separation. A statistical estimate of drug dissociation is needed as the time scale for separation is large compared to that which can be accessed by simulations. These probabilities can be converted to give the equilibrium drug binding constant to compare with observations. We have used our method to study the binding of two different drugs. One is an antitumour drug -- daunomycin which is widely used clinically. This drug is an intercalator that binds to DNA by insertion of its planar structure into the space between DNA base pairs. Another drug studied is a minor groove binding antibiotic drug -- netropsin. This drug is considered as the paradigm of minor groove binders although it is too toxic for clinical use. Our calculated binding constant is in good agreement with observations for both drugs. The temperature dependence of the binding constant at premelting temperatures is found to obey the van't Hoff relation in agreement with experiments. Our method has also been used to study the effect of drug binding on the dynamical stability of DNA base pairs at the binding site. Such a study can give insight into how drugs hinder the transcription and replication processes. Our method is applicable to other molecule-ligand binding systems such as DNA-carcinogen, DNA-protein and protein-ligand complexes. ^* In collaboration with Professor E. W. Prohofsky. Work supported in part by ONR grant N00014-92-K-1232.